A fascinating post by Ron Clutz shows how so-called “skeptics” of global warming misuse data, and how they show it.
He discusses sea surface temperature, titling his post “Global Ocean Cooling in September,” and it appears to be categorized under both “Oceans Make Climate” and “Temperature Trend Analysis.” The main thrust is portrayed in this graph:
It shows sea surface temperature according to the HadSST3 data set since January of 2015, not just for the globe as whole, but for each hemisphere, and for the tropics. I’ll focus on the global data, which I’ll plot in a way I think is much clearer:
I’ve also included a straight-line fit to the data, which Ron Clutz didn’t, in spite of the fact that he categorized his post under “Temperature Trend Analysis.”
In fact Clutz talks a lot about how sea surface temperature cooled during September of 2017 in all four regions, and emphasizes how large the decline was. But this isn’t trend analysis. It might be called fluctuation analysis although I think calling it any kind of “analysis” is too generous.
We can go a bit further by showing a likely range in which the trend lies. I’ll do that here with dashed blue lines above and below the estimated linear trend, and I’ll add a red line to show how it compares to the pre-existing trend:
Two things are worth noting. First, the likely range of where the trend really lies is large. That’s because when you estimate a trend with such a short span of time, the uncertainty in your estimate is huge. Second, the pre-existing trend line falls within that large range.
What pre-existing trend? This one:
This shows HadSST3 data since 1975. Note that the pre-existing trend is beyond doubt. A downward “trend” since 2015 is closer to beyond belief, especially since it’s the kind of persistent up-and-down fluctuation that has been happening, without cease, all along. To take a fluctuation which is absolutely nothing out of the ordinary, ignore any kind of actual trend analysis, then classify your post “Temperature Trend Analysis,” seems quite misleading to me.
It’s even better if we come up with some estimate of the level of uncertainty in the two trend rates — the pre-existing trend and the “since 2015” trend. Here you are:
Perhaps if he ever reads this post, he’ll protest that it was just about ocean cooling in September, simply stating facts and showing data. As for the lack of trend analysis, maybe putting it in that category was just a oversight!
My opinion: the whole thing was his attempt to discredit man-made climate change by suggesting that sea surface temperature is trending down. Maybe he even believes that it is — but in that case, my opinion is that he desperately needs either remedial instruction in trend analysis, or any instruction at all.
It’s particularly hard to believe proclamations of purely innocent posting of data and graphs when the “Summary” at the end begins with this:
We have seen lots of claims about the temperature records for 2016 and 2015 proving dangerous man made warming. At least one senator stated that in a confirmation hearing. Yet HadSST3 data for the last two years show how obvious is the ocean’s governing of global average temperatures.
Ron Clutz has taken a less-than-three-year period of a single data set and used it to imply that the idea of “dangerous man made warming” is bunk.
I suppose in the grand scheme of things, Ron Clutz and his blog musings are negligible. What’s really important to learn from his post is how climate deniers can not only come to the wrong conclusion, but how they can present it in a way which is persuasive if you don’t look at it with a critical eye. One of the main techniques, one which happens again and again and again, is to draw sweeping conclusions based on a time span which is way too short for the job.
I strongly suspect that Ron Clutz isn’t competent to do a real trend analysis, or he wouldn’t be making this kind of post. If he is competent at trend analysis, if he does know that the last few years show the same kind of fluctuation we’ve been seeing all along with no serious departure from the trend, then I would suggest that the misleading aspects of his post aren’t innocent.
This blog is made possible by readers like you; join others by donating at My Wee Dragon.